Academic
Publications
Detecting Pedestrians Using Patterns of Motion and Appearance

Detecting Pedestrians Using Patterns of Motion and Appearance,10.1109/ICCV.2003.1238422,Paul A. Viola,Michael J. Jones,Daniel Snow

Detecting Pedestrians Using Patterns of Motion and Appearance   (Citations: 584)
BibTex | RIS | RefWorks Download
This paper describes a pedestrian detection system that in- tegrates image intensity information with motion informa- tion. We use a detection style algorithm that scans a detec- tor over two consecutive frames of a video sequence. The detector is trained (using AdaBoost) to take advantage of both motion and appearance information to detect a walk- ing person. Past approaches have built detectors based on motion information or detectors based on appearance in- formation, but ours is the first to combine both sources of information in a single detector. The implementation de- scribed runs at about 4 frames/second, detects pedestrians at very small scales (as small as 20x15 pixels), and has a very low false positive rate. Our approach builds on the detection work of Viola and Jones. Novel contributions of this paper include: i) devel- opment of a representation of image motion which is ex- tremely efficient, and ii) implementation of a state of the art pedestrian detection system which operates on low res- olution images under difficult conditions (such as rain and snow).
Cumulative Annual
    • ...The method is very adaptive to dynamic environments [21]...
    • ...The system consists of three modules; the detecting and tracking algorithm [21], the model of possible route estimation and the sound and optical signals...
    • ...Detection and tracking of pedestrians are based on an enhanced implementation of the algorithm proposed in [21] that incorporates both appearance and motion information in near real-time...
    • ...The detection and tracking algorithm originates from the pedestrian detection system proposed by Viola, Jones, and Snow [21] that incorporates both appearance and motion information in near real-time...
    • ...For instance, optical flow possibly provides more motion information than difference images, but it is prohibitively expensive to compute [21]...

    Ioakeim G. Georgoudaset al. An Anticipative Crowd Management System Preventing Clogging in Exits D...

    • ...Recent progress in human detection has advanced the frontiers of this problem in many aspects, e.g., features, classifiers, testing speed, and occlusion handling [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11]...
    • ...Various features have been applied to detect pedestrians, e.g., Haar features [7] and edgelet [10]...
    • ...Cascade (e.g., [7], [11]) and integral image (e.g., [14], [8]) were widely used to accelerate human detection...

    Jianxin Wuet al. Real-time human detection using contour cues

    • ...Several 2-D features (i.e., derived from image intensities) were proposed for vehicle and pedestrian detection, e.g., Haarlike features [10]–[12] and histograms of oriented gradients [13], [14]...

    Ping-Han Leeet al. Viewpoint-Independent Object Detection Based on Two-Dimensional Contou...

    • ...Based on this, Viola et al. [3] later proposed the AdaBoost cascade, a strong classifier composed by layers of threshold-rule weak classifiers, to exploit a variation of HWs, called Haar-like features...

    Guilherme V. Carvalhoet al. A weighted image reconstruction based on PCA for pedestrian detection

    • ...In [3], [13] and [4] algorithms for pedestrian detection and tracking are proposed and in [15] an overview on vision-based pedestrian detection for intelligent vehicles is provided...

    Tohid Ardeshiriet al. Bicycle tracking using ellipse extraction

Sort by: